Is Magnesium a marker of disordered mineral metabolism in males with idiopathic recurrent calcium urolithiasis¿ Observations

Magnesium Research. Volume 16, Numéro 3, 192-205, September 2003, ORIGINAL ARTICLE

Auteur(s) : Angelika Schmiedl, Paul Otto Schwille , Mineral Metabolism and Endocrine Research Laboratory, Department of Surgery and Urology, University of Erlangen, Germany .

Illustrations

ARTICLE

Auteur(s) : Angelika Schmiedl, Paul Otto Schwille

Mineral Metabolism and Endocrine Research Laboratory, Department of Surgery and Urology, University of Erlangen, Germany

Introduction

Abbreviations: IRCU – idiopathic recurrent calcium urolithiasis; Mg – magnesium; Ca – calcium, Pi – inorganic phosphate; N-Alb-P – non-albumin protein; [N-Alb-P] – non-albumin protein concentration; FE – fractional excretion; MA – metabolic activity of IRCU

²Magnesium (Mg) is the fourth most common cation in the human body, inside cells it is second, after potassium. Mg in the extracellular space represents less than 1% of Mg reserves, and in serum Mg it is tightly controlled by the kidney [1]. Low Mg status in the rat is related to nephrocalcinosis [2], in humans to hypertension, cardiovascular diseases and diabetes mellitus type 2 (for overview see ref. 3, 4), all known for their proneness to pathological calcium phosphate (CaPi) calcifications. Mg can modulate calcium (Ca) oxalate crystallization in several regards (for details see ref. 5), and Mg treatment of patients with idiopathic recurrent Ca urolithiasis (IRCU), mainly impressing as Ca oxalate lithiasis, was found effective in stone metaphylaxis [6]. Interestingly, the state of Mg in urine, plasma and cells has received little attention in the research of IRCU pathophysiology, leaving unanswered whether this ion plays a role. In earlier reports urinary Mg was normal in stone patients [7, 8], while Mg intake appeared low [7].

Regarding IRCU etiology, the generally held view is that Ca oxalate crystals form in the thin parts of the nephron, grow and adhere to epithelium, and grow further until crystal aggregates and microliths form, ending up in enhancement of crystalluria and stones [9, 10]. However, crystals and stones not only contain Ca and oxalate but also proteins [11, 12], and, as shown by a sensitive analytical procedure, all stones contain Pi [13]. Further, in IRCU males the most common urine crystal is isotropic (by petrographic microscopy) amorphous CaPi [14], in macromolecule-free liquids known to form at a millimolar ratio of 0.7 [15]. Existing equations and physico-chemical equilibria of Ca oxalate and CaPi, when used for assessing the crystallization risk (upper limit of metastability of these stone phases) in urine, do not take into account the electrostatically charged ionic sites of urinary proteins (albumin, non-albumin), enabling these to attract Ca and Pi ions, thereby serving as crystallization seed; this leaves unanswered whether in protein-containing urine Ca/Pi < 0.7 is compatible with stone formation. Also, when IRCU is evaluated as a whole, i.e., neglecting differences in Ca-uria, the calculated CaPi and calcium oxalate crystallization risk is not different from controls, contrasting with the accompanying highly different CaPi crystalluria [14]. It follows that, relative to changes of urinary oxalate and Ca oxalate crystalluria, the combined changes of urinary Ca, Pi and proteins might play a dominant role in IRCU; if this is so, the associated Mg status is unknown.

Low fasting Mg-uria signals that the body is in need of Mg conservation by the kidney, a state detectable before any decline of extracellular Mg [1]. We, when screening the laboratory data and clinical severity of stone disease [in the following termed metabolic activity (MA)] of IRCU patients, were impressed by the fact that when fasting Mg-uria was low, then urinary pH, citrate, sodium, Ca, potassium, Ca/Pi, proteins and MA were low too, but plasma levels of glucose and insulin were higher-than-normal, and vice versa. If true, this spectrum would be difficult to reconcile with Mg itself as a small-molecular inhibitor of CaPi and Ca oxalate crystallization and stone formation (see ref. 5, 6), but rather a role of Mg as a marker of disturbed metabolism, especially minerals; these latter might modulate stone-forming processes. From the screening of data the impression was also that while urinary oxalate, supersaturation of Ca oxalate and acid Ca phosphate (brushite) appeared unsuspicious, MA (conceived as an integral indicator of all crystal — and stone — generating and inhibiting processes [16] still rises.

In the present work we report more details as to whether fasting Mg-uria, alone or in combination with other variables, impacts on urinary Ca/Pi and MA, when these are set as outcome measures. The Mg/potassium ratio, glycemia and insulinemia were also evaluated, because the former in plasma of patients with cardiovascular disorders is increasingly considered as an indicator of “sick cell disease” [17], whereas the latter two might be elevated when Mg status is low [18]. The results suggest that understanding of renal Ca stone pathophysiology in IRCU benefits from knowledge of the status of Mg, the associated urinary Ca/Pi ratio, and other minerals and substances in fasting urine and plasma.

Material and Methods

Patients

All were white male European residents of the North Bavarian area in Germany, having experienced more than one stone episode in the past. IRCU diagnosis was made by history, KUB X-ray, and stone analysis (X-ray diffractometry, polarization microscopy, wet-chemical analysis), the latter documenting that stones contain Ca and oxalate, in approx. 30% also CaPi salts. In the remaining patients it was not specifically studied whether there is admixture of CaPi to stones; instead – during an in vitro Ca tolerance test – the capability of urine to accumulate not only Ca but also Pi was examined, and a crystal/urine molar Pi ratio ≥ 1.0 was regularly found [19]. A total of 284 adult male patients could be recruited. Excluded were patients with extra-renal Mg loss and enteric hyperoxaluria (e.g. gastrointestinal resections), urinary tract infection (bacillus proteus and others), hematuria (dipstick-positive urine), systemic disorders (oxalosis, overt forms of diabetes mellitus, pHPT, RTA), cystitis and prostatitis, the latter two to minimize that proteinuria was caused by them. A plain KUB X-ray film, obtained shortly before admission to the hospital, was interpreted by two independent observers, in conjunction with disease history serving as the basis for assessment of MA (see below). All patients denied anti-stone medication, or vitamin and mineral supplementation of daily food during the past 6 weeks, and all were advised to minimize intake of food with known high oxalate and “salt” (sodium chloride) content; fluid intake during the 12-15 h night period preceding the ambulatory laboratory examination was restricted to tap water. After detailed instruction about the study goals, all agreed to the investigations, carried out according to the principles of the Declaration of Helsinki.

Laboratory examination

The standardized examination protocol [20] formed the basis for obtaining the data reported herein. In brief: collection of a 24 h urine while eating an unrestricted home diet (see above), ambulatory presentation in the laboratory, measurement of mean blood pressure (twice in a recumbent position at the non-dominant forearm), stimulation of diuresis by drinking of 300 ml distilled water to achieve approx. 1 ml urine flow per min, bladder voiding, withdrawal of blood (without stasis from a pre-warmed forearm vein) into pre-chilled heparinized tubes and a 10 ml syringe [for preparation of plasma ultrafiltrate, using 10 kD pore size cellulose nitrate membrane (Sartorius, Göttingen, Germany) and N₂ pressure], urine collection from a timed (2 h) period. Aliquots of plasma and paper-filtered (Whatman No. 3) urine were prepared, and either immediately analyzed or stored at – 80°C.

Study design

A cross-sectional trial was organized (for age and other anthropometric features see tables 2 and 1). The overlap of patients in present and previous work [16] was about 50%. Tertiles of fasting urinary Mg excretion (low: I, considered as “referent”; intermediate: II; high: III) were set up, expecting that this design together with the sample size allows the detection of abnormal Mg-emia (note that despite Mg deficiency normal Mg in serum [1] and renal tissue [2] were observed) and other abnormalities in blood, plasma and urine, especially whether there are variations of urinary supersaturation, Ca/Pi and Mg/potassium. To avoid possible bias in achieving these goals, the ratio of patients with normo- or idiopathic hypercalciuria, and renal stones present or absent at the time of examination, was kept roughly equal among Mg-uria tertiles.

Analyses

Among the methods used were: total calcium in plasma, ultrafiltrate and urine (by complexometry), ionized Ca in plasma ultrafiltrate [by a colorimetric method [21]], Pi [by a colorimetric micromethod [22]], urinary pH (by glass electrode), plasma glucose (enzymatically) and insulin [which is known to stimulate renal sodium retention and, together with fasting plasma glucose, is accepted as a surrogate marker of insulin resistance of peripheral organs [23] (by in-house radioimmunassay)]; Mg in urine and plasma (by atomic absorption spectrophotometry). Previously communicated were details of the measurement of the bone crosslinks deoxypyridinium and pyridinium [indicating bone resorption (by high performance liquid chromatography [16]], total urinary protein and albumin [16], and plasma oxalate [24]. For intact plasma parathyroid hormone a commercial kit was used (Nichols Institute Diagnostika, Bad Nauheim, Germany), for all other analyses, including urinary oxalate, urea-nitrogen (an indicator of protein intake), and those required for the estimation of urine supersaturation (see below), routine and well-established methods were used.

Calculations, statistics

Conventional formulas were used [urinary total clearances and FE (for Mg and Ca using the plasma ultrafiltrable portion)]. Urinary non-albumin protein (N-Alb-P) excretion was taken as the difference between total protein and albumin. Hydroxyapatite, brushite and Ca oxalate supersaturation of urine was calculated using software EQUIL-2 [25], and expressed as free energy (DG). MA was scored on the basis of clinical severity, forming 5 different degrees of activity of stone formation during the 2-year period preceding the laboratory examination. In brief: group 1 – lumbar pain but no stone present; group 2 – number of new stones identifiable within the kidney; group 3 – stone growth, as assumed from the greatest diameter of the same stone(s) in a previous comparable X-ray; group 4 – documented spontaneous stone passage(s); group 5 – stone removal(s) by surgery or equivalent procedure. In stone-free situations (group 1; healthy individuals, see table 1) MA was set 1.0; to obtain MA from the other groups, these were further weighted by multiplication (factor 2, group 2; 5, group 3; 10, group 4; 15, group 5), then summed. In several instances, including MA, log₁₀ transformation of numerical values led to symmetric distribution, allowing parametric statistical tests. Results are given as mean (SE) or mean and range, as appropriate. One way ANOVA and post-hoc test (Scheffé), and Chi-square test were used; the level of significance of differences was taken as p ≤ 0.05. Simple correlations (Pearson), bi- and multivariate logistic regression analysis were calculated, using STATISTICA software (Statsoft, Tulsa, OK; USA).

Table 1. Characteristics of all IRCU patients as studied during fasting (the total number is 284, except where indicated in brackets). For abbreviations see text

161	< 3.3
Ca/Pi; mM/mM	0.41	0.05	4.0	0.03-0.17**
Mg/Potassium; µM/mM	37	6	163	30-70***

⁺: limits of normalcy and ranges observed in similarly aged adult males in the author's laboratory (see also ref. 8, 20, 46); *, **, ***: mean values are 2.9, 0.12 and 46, respectively.

Results

Stone patients as a whole

According to table 1, giving an overview on parameters of possible relevance for IRCU pathophysiology, the mean, minimum and maximum values of several of these deviate markedly from the respective limits as observed in normals of similar age: apparently higher were body mass index and mean blood pressure, fasting urinary pH and excretion of volume, citrate and proteins, the ratios Ca/Pi and Mg/Potassium, glucose; apparently lower were plasma Pi and daily urinary excretion of Mg. Urinary (daily and fasting) oxalate and urea-nitrogen appeared unsuspicious. In fasting blood bicarbonate appeared low, pH normal, insulin high. This spectrum of abnormalities supports the idea that IRCU includes subgroups characterized by low or high levels of Mg-uria, urinary Ca/Pi, Mg/Potassium, citrate and volume, plasma insulin and glucose.

Stone patients with different fasting Mg-uria

The splitting of IRCU according to Mg-uria tertiles led to the data listed in table 2. The three subsets were similarly aged, but MA was elevated in tertiles II and III vs I (= referent); body weight – not body mass index – was high in tertile I, low in III. In daily urine of tertile I Mg and oxalate were lowest, Ca highest, while sodium, other variables and Ca oxalate supersaturation (data not shown) were comparable.

Table 2. Mg excretion rate by tertiles (totaling to 284 patients) and associated variables in urine, blood and plasma. Mean values (SE), except MA, fasting urinary Mg and protein excretion, fasting blood bicarbonate and pH [mean (range of values)]. ⁺: except [ ]; NC/I-HC: normocalciuria/idiopathic hypercalciuria. *: statistically indistinguishable from I (x²-test); ¹⁾: statistics based on decadic log of numerical values. ^A: F-value significant. For other informations see text and table 1

	Tertiles
	I	II	III
Number of patients⁺	94	95	95
Mg in fasting urine^A; µM	150 (50-192)	246^c (196-296)	383^{c; f}(296-792)
General features
Age; y	43.5 (1.1)	42.8 (1.2)	43.2 (1.3)
Weight^A; kg	83.9 (1.0)	82.5 (1.0)	79.5^{c; d} (1.0)
Body mass index; kg/(m)²	26.9 (0.3)	26.5 (0.4)	26.1 (0.4)
Mean blood pressure; mm Hg	107 (2) [82]	109 (2) [75]	108 (2) [76]
Stones; present/absent	38/56	47/48*	47/48*
NC/I-HC	69/25	60/35*	65/30*
MA^A; score¹⁾	33 (1-230)	45^a (1-271)	39^a (1-173)
Variables in 24 h urine
Sodium; mM	190 (8)	179 (8)	178 (7)
Calcium^A; mM	5.8 (0.3)	4.9^c (0.2)	5.5^d (0.3)
Pi; mM	30 (0.9)	29 (0.8)	30 (0.8)
Potassium; mM	65 (3)	62 (2)	61 (4)
Mg^A; mM	3.7 (0.1)	4.0 (0.1)	4.6^{c; f} (0.1)
Urea-Nitrogen; mM	0.79 (0.03)	0.77 (0.03)	0.77 (0.03)
Oxalate^A; mM**	0.25 (0.01) [31]	0.26 (0.02) [26]	0.30^{a; d} (0.02) [21]
Variables in 2 h fasting urine
Volume^A; ml	177 (12)	225^b (15)	280^{c; e} (17)
Creatinine clearance^A; ml/min	113 (4)	115 (3)	128^{b; e} (4)
Potassium^A; mM	7.3 (0.3)	7.9 (0.3)	8.7^{b; d} (0.4)
Mg/Potassium^A; µM/mM	24 (1)	37^a (2)	51^{a; b} (2)
Ca^A; µM	235 (10)	300^c (11)	425^{c; f} (20)
Pi^A; mM	1.23 (0.06)	1.23 (0.06)	1.06^{a; d} (0.06)
Desoxypyridinium^A; nmol	6.7 (0.4) [54]	8.1^a (0.7) [54]	9.8^a (2.0) [62]
Total protein^A; mg¹⁾	7 (0.8-171)	9^b (0.9-161)	9^c (0.8-101)
Albumin^A; mg¹⁾	0.77 (0.05-9.8) [64]	2.4^a (0.08-57) [60]	2.1^a (0.09-30) [61]
N-Alb-P^A; mg¹⁾	4.1 (0-13) [64]	7.9^b (0-160) [60]	6.8^c (0-100) [61]
Urea-nitrogen^A; µM	62 (2)	74^a (5)	75^c (3)
Oxalate; mg	24 (2) [89]	23 (2) [91]	23 (1) [93]
Hydroxypatite^A; DG***	2.1 (0.3) [75]	2.7 (0.4) [72]	3.7^{c; f} (0.3) [80]
Calcium oxalate; DG	0.9 (0.1) [75]	0.8 (0.1) [72]	0.9 (0.1) [80]
Fasting blood
Bicarbonate^A; mM/l	23.2 (18.0-29.8) [93]	23.6 (11.9-30.6) [90]	23.9^a (18.7-31.0) [93]
pH^A	7.40 (7.36-7.46) [93]	7.41 (7.32-7.46) [90]	7.41^a (7.36-7.49) [93]

**: note that the means of Ca oxalate supersaturation ( ΔG) in tertiles were statistically unchanged; ***: note that the means of D G brushite and D G octacalcium-phosphate supersaturation were negative (indicating dissolution)
^a: ≤ 0.05; ^b: p < 0.01; ^c: p < 0.001 vs I; ^d: p ≤ 0.05; ^e: p < 0.01; ^f: p < 0.001 vs II.

Table 2 (followed). Mg excretion rate by tertiles (totaling to 284 patients) and associated variables in urine, blood and plasma. Mean values (SE), except MA, fasting urinary Mg and protein excretion, fasting blood bicarbonate and pH [mean (range of values)]. ⁺: except [ ]; NC/I-HC: normocalciuria/idiopathic hypercalciuria. *: statistically indistinguishable from I (x²-test); ¹⁾: statistics based on decadic log of numerical values. ^A: F-value significant. For other informations see text and table 1

	Tertiles
	I	II	III
Fasting plasma
Total Mg^A; mM/l	0.83 (0.01)	0.85^a (0)	0.86^c; (0)
Ultrafiltrable Mg; mM/l	0.65 (0.01)	0.66 (0)	0.67 (0)
Potassium; mM/l	4.20 (0.05) [33]	4.25 (0.04) [34]	4.20 (0.05)
Mg/Potassium; mM/mM	0.20 (0.00) [33]	0.21 (0.00) [34]	0.21 (0.00) [43]
Total Ca; mM/l	2.34 (0.01) [87]	2.35 (0.01) [86]	2.34 (0.01) [88]
Ultrafiltrable Ca; mM/l	1.48 (0.01)	1.48 (0.01)	1.48 (0.01)
Ionized Ca^A; mM/l	1.20 (0.01) [87]	1.20 (0.01) [86]	1.18^d (0.03) [88]
Parathyroid hormone^A; pM/l	2.6 (0.1) [71]	2.8 (0.1) [65]	2.3^b; e (0.1) [70]
Pi; mM/l	0.95 (0.02)	0.97 (0.02)	0.94 (0.01)
Sodium; mM/l	142 (0.3)	143 (0.3)	143 (0.3)
Glucose^A; mM/l	5.0 (0.06) [84]	4.9 (0.06) [85]	4.7^b; e (0.06) [89]
Insulin^A; pM/l	136 (14) [85]	122 (7) [85]	93^c; e (7) [89]
Oxalate^A; µM/l	1.71 (0.07) [41]	1.78 (0.1) [36]	1.95^a (0.09) [29]
Fractional clearances in fasting urine
Mg^A;%	1.9 (0.07)	2.8^c (0.08)	3.8^c; f (0.1)
Sodium^A;%	0.61 (0.03)	0.71^a (0.04)	0.79^c (0.04)
Ca^A;%	1.3 (0.06)	1.5^a (0.07)	2.0^c; f (0.1)
Pi;%	8.4 (0.4)	9.2 (0.5)	9.4 (0.6)
Potassium;%	14 (0.6)	14 (0.5)	13 (0.5)
Oxalate;%	119 (8) [37]	106 (7) [31]	112 (9) [27]

**: note that the means of D Ca oxalate supersaturation were unchanged; ***: note that the means of D G brushite and D G octacalcium-phosphate supersaturation were negative (indicating dissolution)
^a: p ≥ 0.05; ^b: p < 0.01; ^c: p < 0.001 vs I; ^d: p ≥ 0.05; ^e: p < 0.01; ^f: p < 0.001 vs II.

In tertile I all individual Mg-uria values were < 4 mg/2 h (in approx. 65% < 2.5 mg/2 h), FE-Mg was also low (range 0.5-3.5%), and the mean plasma total Mg was lowest; in tertile II the range of fasting Mg excretion (4.7-7.1 mg/2 h) was close to the one observed previously in healthy controls [4.6-9.8 mg/2 h [8]]. In contrast, in tertile III, showing a mean FE-Mg of 3.8% (corresponding to the upper limit of normalcy; table 1), the associated fasting Mg excretion of numerous patients was high, as was plasma total Mg. This indicates that, according to Mg in daily urine, Mg-uria and Mg-emia during fasting, IRCU comprises subsets with low (tertile I), normal (tertile II), and undecided (tertile III) Mg status.

Figure 1 shows that along with the increase of fasting Mg-uria in tertiles (table 2) there was also an increase of urinary Ca/Pi, sodium, citrate and the bone collagen crosslink deoxy-pyridinium [the mean values for the crosslink pyridinium were 28, 32, 39 nmol per 2 h, for tertiles I, II and III, respectively, differing significantly in II and III from I]. Plasma parathyroid hormone, unchanged in tertile II vs I, was decreased in III, accompanied by slightly elevated urinary pH. In 156 of the 284 patients (≈ 55%) the Ca/Pi ratio was > 0.25 (assuming that this figure safely indicates the upper limit in normals; see also table 1), but a Ca/Pi ratio ≥ 0.7 [the reported minimum for amorphous CaPi precipitation in protein-free solutions [15]] was present in only 4 patients of tertile I (approx. 1%), 7 (2%) in II, 23 (8%) in III.

Table 2 additionally shows that increasing fasting Mg-uria was accompanied not only by increasing urinary Ca/Pi (figure 1), but also urinary (not plasma) Mg/Potassium ratio, excretion of protein (albumin, N-Alb-P), urea-nitrogen, and DG hydroxyapatite. DG Ca oxalate did not differ among tertiles. It is noteworthy that the increase of urinary Ca/Pi and DG hydroxyapatite in tertiles II and III vs I was due to a strong rise of urinary Ca vis-à-vis decreasing Pi excretion, the increase of Mg/Potassium due to a strong rise of Mg (126%) vis-à-vis the only marginal increase of potassium (18%). There also were higher (in tertiles II and III vs I) blood bicarbonate and pH (albeit the latter was within the normal range), and there was a rise of urinary volume, creatinine clearance (in this work considered equivalent to glomerular filtration rate), FE of Mg, Ca, sodium, but not of FE Pi, potassium and oxalate; significantly lower than in I were glycemia and insulinemia. Collectively (from the data in table 2 and figure 1), tertile I patients not only exhibit a low Mg status in combination with low citraturia and a tendency toward metabolic acidosis, but also signs of type 2 diabetes mellitus; conversely, the tertile III patients show a tendency toward urinary loss of sodium, Ca, Mg and proteins, in combination with an increase of MA but attenuation of insulinemia, glycemia and acidosis.

Interrelationships

From a large correlation matrix, encompassing the numerical and log data of the variables as obtained during fasting (see figure 1 and table 2), information was sought as to possible effectors of urinary Ca/Pi and MA. In a first step significant (p ≤ 0.05) correlations of Ca/Pi with other variables were identified, thereafter the same variables were contrasted with MA (bivariate regression analysis). Also included were urinary N-Alb-P concentration ([N-Alb-P]), log Mg/Potassium, citrate excretion, Ca oxalate and hydroxyapatite supersaturation, because all these, when correlating merely insignificantly with Ca/Pi, correlated significantly with log MA, and vice versa.

According to table 3, significant correlates of urinary Ca/Pi were the excretion rate (or/and FE) of Mg, Ca, sodium, the two supersaturation products, urinary Mg/Potassium and MA (all positive), and urinary Pi (negative); significant correlates of MA were urinary sodium (excretion rate and FE), FE Mg, [N-Alb-P] and, weakly, Ca/Pi; Mg and citrate excretion were only borderline significant. The association of oxalate excretion and FE oxalate with either Ca/Pi or MA was insignificant (data not shown). Upon multivariate regression analysis (table 4) the model best fitted to urinary Ca/Pi revealed as significant predictors urinary Ca excretion and Mg/Potassium ratio (positive), and urinary Pi excretion (negative), with citrate excretion being only borderline significant; together, the four variables explain approx. 85% of the total variation of urinary Ca/Pi. Conversely, the model best fitted to MA revealed urinary [N-Alb-P], Mg/potassium ratio and sodium excretion as significant positive predictors; together the three variables explain approx. 6% of the total variation of MA. Therefore, despite the slightly significant correlation of MA and urinary Ca/Pi ratio (see table 3), > 90% of the variation of MA was contributed by factors other than Ca/Pi (presumably individual proteins among the N-Alb fraction of urinary proteins; see below). Neither oxalate excretion nor DG Ca oxalate were effectors of MA.

Table 3. Bivariate correlations (r: coefficient; p: significance level) of log Ca/Pi in urine (U) and log MA with other variables

	U-Ca/U-Pi		MA
	r¹⁾	p	r¹⁾	p
U-Mg²⁾	0.40	< 0.001	0.11	0.07
FE-Mg	0.29	< 0.001	0.17	0.003
U-Sodium	0.29	< 0.001	0.13	0.03
FE-Sodium	0.19	0.002	0.16	0.008
U-Ca	0.60	< 0.001	0.07	0.26
FE-Ca	0.48	< 0.001	0.09	0.14
U-Pi	-0.07	< 0.006	-0.09	0.13
U-log Ca/Pi	–	–	0.12	0.04
U-log Mg/Potassium	0.30	< 0.001	0.09	0.13
U-Citrate	0.19	0.003	0.12	0.06
U-[N-Alb-P]	0.01	0.93	0.19	0.008
U-Hydroxyapatite	0.27	0.001	-0.02	0.75
U-Ca oxalate	0.31	0.001	-0.09	0.90

¹⁾: n = 284 paired observations, except urinary [N-Alb-P] (n = 191), hydroxyapatite (n = 273), Ca oxalate (n = 227), and citrate excretion (n = 243). ²⁾: for the dimension of variables, other abbreviations and information see tables 1, 2 and text.

Table 4. Determinants of urinary (U) log Ca/Pi and log MA according to multivariate regression analysis. Beta: partial correlation coefficient; SE: standard error of Beta; p: significance level; R²: coefficient of the model (adjusted for confounders). For other information see table 3


R² 0.85, n = 243, p < 10^-5

Discussion

IRCU as a model

According to presented data, a rise of Ca/Pi ratio in urine, to a certain extent of MA (synonymous the clinical severity of Ca stone disease), is paralleled by a rise in Mg-uria, natriuresis, proteinuria and several other abnormalities, whereas an impact of urinary oxalate (excretion, FE) or Ca oxalate supersaturation on Ca/Pi and MA is not demonstrable. The rise of plasma oxalate in tertile III vs I (table 2) is not explicable by available data, but may indicate some degree of oxalate overproduction or renal retention during fasting. These findings are new, allowing us to postulate that in IRCU a primacy of disordered urinary minerals, but not oxalate, may exist, and that Mg most likely plays a marker role. In fact, there is the impression that during a low Mg status (inferred from low Mg in urine and plasma) malregulation of mineral and acid-base status, insulin and glucose metabolism are frequent and possibly interdependent. When modeling renal Ca stone disease on the basis of metabolic events (e.g., the response to variation of mineral nutrition), Mg and Ca appeared to be of greater importance than oxalate [26]. For example, dietary Mg deficiency increases Pi-uria independent of parathyroid gland function [27], and even short-term Mg deficiency causes tissue damage, modification of protein synthesis and ion fluxes [28]. In earlier work we found that in stone patients vis-à-vis controls the energy as supplied with macronutrients is too high, although there was no predominance of protein (especially animal protein), while the supply of Pi and Ca may be low, of oxalate normal [29]; however, in that work Mg was not tested. Therefore, intake of food with anomalous composition may be ruled out, except perhaps in low intake of Mg. Hence, the metabolic state of the patients as observed during fasting may reflect variations of intrinsic abnormalities, related to Mg status alone or additional to yet unknown factors. Others postulated that a fundamentally and probably genetically disturbed homeostasis of cellular and extracellular minerals and proteins, including transporters of ions across cell membranes, underlies IRCU [30]. Clearly, to investigate such details was beyond the scope of our work.

Mg status of IRCU – Links to sodium

Mg deficiency, in Germany amounting to approx. 30% of healthy adults, has been inferred from hypo-Mg-emia [31] that is presumably due to insufficient dietary intake of Mg. The present work shows that, when the home diet was unrestricted with respect to Mg, lower-than-normal fasting Mg-uria (excretion, FE) is characteristic for at least one-third of middle-aged patients, and that the finding co-incides with slightly reduced plasma total Mg but unchanged plasma free (synonymous ultrafiltrable) Mg (tertile I; table 2). Similarly, in previous and preliminary work on cellular minerals, we found reduced total Mg [8], but normal free ultrafiltrable Mg (unpublished data). Thus, similar to healthy individuals [32, 33], at least in some subset of IRCU patients, an incipient Mg deficit may exist and be able to drive the kidney to reclaim Mg, thereby preserving normal free plasma and cellular Mg, but at the expense of the protein-bound Mg fraction. This Mg fraction may indeed play a crucial role, as in tertile III plasma total, but not free, Mg increases when bone efflux of Mg may be present (see below), i.e., plasma macromolecules may buffer a threat exerted by relocation of endogenous Mg. Coincidence of high fasting insulinemia and high glycemia (glucose > 80 mg/dl points toward incipient type 2 diabetes mellitus [34, 35] is ascribed to resistance of peripheral organs to insulin-mediated glucose utilization, and both signs are widely considered as surrogate markers of arteriosclerosis [3]. Insulin resistance is a putative Mg-dependent abnormality and is probably frequent in IRCU [36, 37]; Mg deficiency [3, 38] and numerous other factors have been proposed as underlying causes, including obesity, metabolic acidosis and Pi deficiency [36]. Mg is an essential cofactor in reactions involving phosphorylation (of glucose and other metabolites) and those utilizing phosphorus bonds; therefore, if in our work Mg deficit accounts for high insulinemia (tertile I; table 2), availability of plasma free ionized Mg and Mg fractions inside cells would have allowed more insight into the gluco-metabolic and ionic events at the levels of cell membranes. Irrespective of the factor(s) causing insulin excess, one of the unwanted effects of this hormone is sodium retention [23], implying that high fasting insulinemia may be associated with low natriuria, and vice versa; in the present work these constellations are impressively demonstrable in tertiles I and III patients (table 2), respectively. Prelimary work of our own revealed that among IRCU patients sodium retention from a test meal is frequent [39]; in agreement with this would be insufficient release of a natriuretic ouabain-like factor [40], low plasma levels of, or impaired renal-tubular sensitivity to, atrial natriuretic peptide [41], factors tending to increase natriemia and to expand extracellular volume. These would have deleterious effects on the maintenance of the interior milieu, regulation of hemodynamics included. Therefore, the constancy of plasma sodium in tertiles (table 2) is interpreted to mean that storage of osmotically inactive sodium in bone and other tissues occurred [42], and that sodium release from bone enhanced natriuresis (see below). Under normal conditions renal sodium and Ca handling largely dictate renal Mg handling [43], while renal sodium and Pi transport are coupled [44]. However, in IRCU the correlations of sodium, Ca and Mg were found to be weaker than in controls [45], and the state of Pi appears low (table 1) [36, 46], while the generation of systemic proton excess may be accentuated owing to hypercaloric (proton-producing) nutrition [29] and Mg deficiency [47]. Therefore, the combination of overweight, urinary loss of the poorly reabsorbable monovalent Pi species [in urine Pi is the physiologically dominant buffer [48]] and acidosis-induced consumption of the base-precursor citrate can explain why in urine of tertile I (relative to III) patients the constellation of low Mg and citrate and high Pi develops, despite the fact that plasma ionized Ca, parathyroid hormone, and natriuresis are within normal limits.

Acidosis, response of bone and kidney – Key to understanding mineraluria, proteinuria and MA

Metabolic acidosis for whatever reason stimulates protein synthesis and induces a variety of chronic diseases, including renal calcifications [49]. Low Mg and low Pi status are known to alter bone minerals [15]. If one assumes that in our work these abnormalities and acidosis exist ab initio, then the observations on the increase of bone resorption, mineraluria and proteinuria in the tertiles II and III patients (table 2, figure 1) become understandable: physico-chemical – not osteoclasts-mediated – dissolution of bone Ca carbonate and other minerals may be operative [15] and upregulates blood bicarbonate and pH [49], but in addition creates a threat to Mg, sodium and other minerals in plasma. Consequently, adaptation of the filtration and excretory function of the kidney via elevation of glomerular filtration rate and diminution of tubular net reabsorption is necessary, enhancing renal elimination of bone-derived sodium, Mg, Ca, potassium and bicarbonate, resulting in a rise of urinary pH. This interpretation, although requiring more direct demonstration, would explain the hyperexcretion of bone resorption markers despite declining plasma parathyroid hormone, the positive correlations of urinary Ca/Pi with Mg and other substances in urine, and with MA (table 3), and the impact of urinary sodium on MA (table 4). The fact that urinary [N-Alb-P] is not a direct predictor of Ca/Pi but is a correlate and positive partial predictor of MA, suggests that in post-glomerular renal tissue low Mg status [28], acidosis [49], or both interfere with protein synthesis (in terms of amount and structure). Alteration of the structure of a given protein may switch its function from inhibition of stone formation to promotion of stones. One example is the loss of phosphorylation – with subsequent loss of inhibitory property – of the multifunctional phosphoprotein osteopontin [50], which is produced by the kidney [51], found in crystals, stones and urine [51], and in the present work is likely to be a member of the urinary N-Alb-P fraction. Another example is the self-aggregation of Tamm-Horsfall protein – normally a stone inhibitor – that switches it to function as a promoter of stones [52]. It remains unclear why net tubular reabsorption of potassium (controlled by distal tubules) and Pi (largely controlled by proximal tubules) remained unchanged in the majority of patients (see FE values in table 2, tertiles II and III). Perhaps the finding reflects that in patients with high MA several defects – recognizable in tertile I – are hidden by activation of bone resorption, counteracting not only extra-osseous acidosis and low Mg but also Pi and potassium status [note that the plasma levels of the latter two are low-normal (tables 1, 2), corroborating the finding that in Europe a low renal Pi threshold is characteristic for IRCU [46, 53], and that in other parts of the world potassium deficiency is a stone risk factor [54]]. Worthy of note, in this setting the renal conservation of Mg was less effective than that of Pi and potassium. In other words: apparent normo-(tertile II) and hyper-(tertile III) magnesiuria may mask that Mg deficiency, undetectable by routine clinical means such as measurement of Mg in daily urine and often non-fasting plasma, is causative and plays an eminent role, at least in the early course of IRCU or during its initiation. This interpretation, needing verification by controlled studies, would explain why it has been questioned that in unclassified IRCU, low Mg status is a stone risk factor [55].

Conclusions

Using examination during fasting, a cross-sectional study design, and reliable analytical techniques, a segment of IRCU patients can be identified as suffering from a combination of low urine and plasma Mg, modest metabolic acidosis, high glycemia and insulinemia. In another subset, exhibiting high Mg-uria and high Mg-emia, dissipation of acidosis and attenuation of glycemia and insulinemia may have been achieved by an increase of parathyroid hormone-independent bone resorption, presumably ending up in an overflow of minerals, alkali and other substances from bone to plasma, eubicarbonatemia and a rise of plasma levels of several minerals. Although in the subset with low Mg, status verification of overt Mg deficiency is not possible with the study design used, it appears that this abnormality, modest metabolic acidosis, high insulinemia and sodium retention are background phenomena in IRCU as a whole. Increase of renal protein synthesis and adaptation of the excretory function of the kidney, necessarily built as defense, elevate urinary protein, Ca/Pi ratio, Mg, pH, and hydroxyapatite supersaturation; together these factors increase the risk of CaPi precipitation and aggravate stone formation at the expense of bone mineral density, a well documented abnormality in IRCU [19]. Because in this environment oxaluria and Ca oxalate supersaturation remain unchanged, the precipitation of Ca oxalate — the major constituent of renal Ca stones — may be secondary to CaPi.

Acknowledgement

We are grateful to K Schwille for technical, ML Rasenack for secretarial assistance. J Wipplinger provided the stone clinical electronic data management.

References

1. Sutton R.A.L., Domrangkitchchaiborn S. Abnormal renal magnesium handling. Min Electrolyte Metab 1993; 19, 232-40.

2. Schmiedl A., Schwille P.O., Bergé B., Markovic M., Dvorak O. Reappraisal of the quantity and nature of renal calcifications and mineral metabolism in the magnesium-deficient rat. Urol Int 1998; 61, 76-85.

3. Durlach J. (1992): Magnesium in der klinischen Praxis; Stuttgart, Jena G. Fischer.

4. Resnick L.M., Altura B.T., Gupta R.K., Laragh J.H., Alderman M.H., Altura B.M. Intracellular and extracellular magnesium depletion in Type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia 1993; 36, 767-70.

5. Schwille P.O., Schmiedl A., Herrmann U., Fan J., Gottlieb D., Manoharan M., Wipplinger J. Magnesium, citrate, magnesium citrate and magnesium-alkali citrate as modulators of calcium oxalate crystallization in urine: observations in patients with recurrent idiopathic calcium urolithiasis. Urol Res 1999; 27, 117-26.

6. Ettinger B., Pak C.Y.C., Citron J.T., Thomas C., Adams-Huet B., Vangessel A. Potassium-magnesium citrate is an effective prophylaxis against recurrent calcium oxalate nephrolithiasis. J Urol 1997; 158, 2069-73.

7. Takeuchi H., Yoshida O. (1989): Magnesium and urinary stones. In: Magnesium in health and disease, eds Y. Itokawa, J. Durlach, pp. 377-82, London, Paris: John Libbey.

8. Schmiedl A., Schwille P.O. Magnesium status in idiopathic calcium urolithiasis – An orientational study in younger males. Eur J Clin Chem Clin Biochem 1996; 34, 393-400.

9. Khan S.R. Tubular cell surface events during nephrolithiasis. Curr Op in Urol 1997; 7, 240-247.

10. Kok D.J. Intratubular crystallization events. World J Urol 1997; 15, 219-28.

11. Doyle I.R., Ryall R.L., Marshall V.R. Inclusion of proteins into calcium oxalate crystals precipitated from human urine: a highly selective phenomenon. Clin Chem 1991; 37, 1589-94.

12. Dussol P., Geider S., Lilova A., Léonetti F., Dupuy P., Daudon M., Berland Y., Dagorn J.C., Verdier J.M. Analysis of the soluble organic matrix of five morphologically different stones. Urol Res 1995; 23, 45-51.

13. Larsson L., Sörbo B., Tiselius H.G., Öhmann S. A method for quantitative wet-chemical analysis of urinary calculi. Clin Chim Acta 1984; 140, 9-13.

14. Herrmann U., Schwille P.O., Kuch P. Crystalluria determined by polarization microscopy. Urol Res 1990; 19, 151-58.

15. Green J. (1996): The role of bone in the maintenance of acid-base homeostasis: Cellular and non-cellular factors. In. Kidney Stones, eds F.L. Coe, M.J. Farus, C.Y.C. Pak, J.H. Parks, G.M. Preminger, pp. 821-49. Philadelphia: Medical and Surgical management Lippincott-Raven Publishers.

16. Manoharan M., Schwille P.O. Oxypurines, protein, glucose and the functional state of blood vasculature are markers of renal recurrent calcium stone-forming processes ? Observations in men with idiopathic recurrent calcium urolithiasis. Clin Chem Lab Med 2002; 40, 266-77.

17. Ryan M.F. The role of magnesium in clinical biochemistry: an overview. Ann Clin Biochem 1991; 28, 19-26.

18. Rosolova H., Mayer jr. O., Reaven G.M. Insulin-mediated glucose disposal is decreased in normal subjects with relatively low plasma magnesium concentration. Metab 2000; 49, 418-20.

19. Schwille P.O., Schmiedl A., Manoharan M. (manuscript in preparation): Calcium oxalate and calcium phosphate nucleation in postprandial urine of males with idiopathic calcium urolithiasis are related to associated free citrate and protein ? An observational study indicating co-precipitation and a modulating role of urine pH.

20. Schwille P.O., Rmenapf G. (1990): Idiopathic calcium urolithiasis – Clinical problems and suggested approaches in an ambulatory stone clinic. In: Renal Tract Stone, eds J.E.A. Wickham, A. Buck, pp 217-238. Edinburgh, London, Melbourne, New York: Metabolic Basis and Clinical Practice – Churchill Livingstone.

21. Robertson W.G., Marshall R.W. Ionized calcium in body fluids. CRC Crit Rev Clin Lab Sci 1981; 15, 85-125.

22. Chen P.S., Toribara T.Y., Warnre H. Microdetermination of phosphorus. Anal Chem 1956; 28, 1756-1760.

23. Solini A., De Fronzo R. Insulin resistance, hypertension and cellular ion transport. Acta Diabetol 1992; 29, 196-200.

24. Manoharan M., Schwille P.O. Measurement of oxalate in human plasma ultrafiltrate by ion chromatography. J Chromatogr B 1997; 700, 261-8.

25. Werness P.G., Brown C.M., Smith L.H. EQUIL-2: a BASIC computer program for the calculation of urinary saturation. J Urol 1985; 134, 1242.

26. Finlayson B., Khan S.R., Hackett R. (1990): Theoretical chemical models of urinary stones. In: Renal Tract Stone, eds J.E.A. Wickham, A.C. Buck, pp. 133-147. Edinburgh, London, Melbourne, New York: Metabolic Basis and Clinical Practice – Churchill Livingstone.

27. Ginn H.E., Shanbour L.L. Phosphaturia in magnesium-deficient rats. Am J Physiol 1967; 212, 1347-50.

28. Petrault I., Zimowska W., Mathieu J., Bayle D., Rock E., Favier A., Rayssiguer Y., Mazur A. Changes in gene expression in rat thymocytes identified by cDNA support the occurrence of oxidative stress in early magnesium deficiency. Biochim Biophys Acta 2002; 1586, 92-8.

29. Scholz D., Schwille P.O., Sigel A. Ernährungsgewohnheiten von Patienten mit Urolithiasis. Fortschr Urol Nephrol 1981; 17, 83-6.

30. Gambaro G., Marchini F., Budakovic A., Checchetto S., Baggio B. Genetic approach to the study of cellular ion transport anomalies in idiopathic calcium nephrolithiasis. Contrib Nephrol 1997; 122, 189-92.

31. Schimatschek H.F., Rempis R. Prevalence of hypomagnesemia in an unselected German population of 16000 individuals. Magnes Res 2001; 14, 283-90.

32. Sutton R.A.L. Renal handling of calcium and magnesium in idiopathic stone disease. Contr Nephrol 1987; 58, 93-100.

33. Barnes B.A., Cope O., Harrison T. Magnesium conservation in the human being on a low magnesium diet. J Clin Invest 1958; 37, 430-40.

34. De Fronzo R.A., Bonadonna R.C., Ferannini E. Pathogenesis of NIDDM. Diabet Care 1992; 15, 318-68.

35. Wälthi M.K., Zimmermann M.B., Spinas G.A., Jacob S., Hurell R.F. Dietary magnesium intake in type 2 diabetes. Eur J Clin Nutr 2002; 56, 409-14.

36. Schwille P.O., Schmiedl A., Herrmann U., Wipplinger J. Postprandial hyperinsulinemia, insulin resistance and inappropriately high phosphaturia are features of younger males with idiopathic calcium urolithiasis; attenuation by ascorbic acid supplementation of a test meal. Urol Res 1997; 25, 49-58.

37. Flessa P., Tosco C., Paganin L., Mioni G., Zattoni F., Guerra U.P. (1999): Insulin sensitivity and arterial blood pressure in stone formers. In: Kidney Stones, eds L. Borghi, T. Meschi, A. Briganti, T. Schianchi, A. Novarini, pp. 603-5. Cosenza: Editoriale Bios.

38. Linda Kao W.H., Folsom A.R., Nieto J., Jing-Ping M.O., Watson R.L., Brancati F.L. Serum and dietary magnesium and the risk for type 2 diabetes mellitus. Arch Int Med 1999; 159, 2151-9.

39. Schmiedl A., Schwille P.O. (1999): Dietary sodium retention and urinary calcium loss in idiopathic calcium urolithiasis (ICU). Correction by a test meal supplemented with alkalinizing magnesium citrate. In: Kidney Stones, eds L. Borghi, T. Meschi, A. Briganti, T. Schianchi, A. Novarini, pp. 561-3. Cosenza: Editoriale Bios.

40. Rivera R., Soldati L., Manunta P., Zerbi S., Adamo D., Bianchi G., Vezzoli G. (1999): Role of plasma circulating ouabain-like factor in idiopathic hypercalciuria. In: Kidney Stones, eds L. Borghi, T. Meschi, A. Briganti, T. Schianchi, A. Novarini, pp. 481-3. Cosenza: Editoriale Bios.

41. Halabe A., Wong N.L., Wong E.F., Sutton R.A. Atrial natriuretic factor levels in renal stone patients with idiopathic hypercalciuria and in healthy controls: The effect of an oral calcium load. Metab 1990; 39, 209-212.

42. Bergstrom W.H., Wallace W.M. Bone as a sodium and potassium reservoir. J Clin Invest 1954; 33, 867-73.

43. Walser M. (1973): Divalent cations: physico-chemical state in glomerular filtrate and urine, and renal excretion. In: Handbook of Physiology, Section 8, eds J. Orloff, R.W. Berliner, pp 555-586. Washington DC: American Physiological Society.

44. Biber J., Murer H., Forster I. The renal type II Na + /Phosphate cotransporter. J Bioenerget Biomembr 1998; 30, 187-94.

45. Schwille P.O., Schlenk I., Samberger N.M., Bornhof C. Fasting urine excretion of magnesium, calcium and sodium in patients with renal calcium stones. Urol Res 1976; 4, 33-43.

46. Schwille P.O., Herrmann U., Schmiedl A., Kissler H., Wipplinger J., Manoharan M. Urinary phosphate excretion in the pathophysiology of idiopathic recurrent calcium urolithiasis: hormonal interactions and lipid metabolism. Urol Res 1997; 25, 417-26.

47. Kinne-Safran E. Renal H + -ATPases. Ann NY Acad Sci 1989; 574, 189-203.

48. Mizgala C.L., Quamme G.A. Renal handling of phosphate. Physiol Rev 1985; 65, 431-66.

49. Alpern R.J., Sakhae K. The clinical spectrum of chronic metabolic acidosis: homeostatic mechanisms produce significant morbidity. Am J Kidney Dis 1997; 29, 291-302.

50. Hunter G.K., Hauschka P.V., Poole A.V., Rosenberg L.C., Goldberg H.A. Nucleation and inhibition of hydroxyapatite formation by mineralized tissue protein. Biochem J 1996; 317, 59-64.

51. Takahiro Y., Sato M., Fujita K., Tozawa K., Nomura S., Kohri K. Effects of citrate on renal stone formation and osteopontin expression in a rat urolithiasis model. Urol Res 2001; 29, 50-56.

52. Hess B., Nakagawa Y., Parks J.H., Coc F.L. Molecular abnormality of Tamm-Horsfall glycoprotein in calcium oxalate nephrolithiasis. Am J Physiol 1991; 260, F 569-F 578.

53. Prié D., Ravery V., Baccon-Gibod L., Friedlander G. Frequency of renal phosphate leak among patients with calcium nephrolithiasis. Kidney Int 2001; 60, 272-6.

54. Tosukowong P., Borronpadungkitti S, Prasongwatana V., Tungsanga P., Jutuporn S., Dissayabutr T., Reungjui S., Sriboonlue P. Urinary citrate excretion in patients with renal stone: roles of leucocyte ATP citrate lyase activity and potassium salts therapy. Clin Chim Acta 2002; 325, 71-8.

55. Ryall R.L., Marshall V.R. (1990): The investigation and management of idiopathic urolithiasis. In: Renal Tract Stone, eds J.E.A. Wickham, A.C. Buck, pp. 307-331. Edinburgh, London, Melbourne, New York: Metabolic Basis and Clinical Practice – Churchill Livingstone.

	U-Ca/U-Pi				MA
	Beta	SE	p		Beta	SE	p
U-Ca	0.56	0.03	< 10^-6	U-[N-Alb-P]	0.18	0.07	0.01
U-Pi	-0.68	0.03	< 10^-6	U-log Mg/Potassium	0.17	0.07	0.02
U-log Mg/Potassium	0.06	0.03	0.002	U-Sodium	0.15	0.07	0.04
U-Citrate	0.05	0.03	0.06	R² 0.06, n = 191, p = 0.0019

	Mean	Minimum	Maximum	Normal⁺
General features
Age; years	43	20.7	70.7	< 60
Body mass index; kg/(m)²	26.5	18.6	55.6	< 25.0
Weight; kg	82	55	130
Height; cm	176	160	200
Mean blood pressure; mm Hg	108 [233]	75	185	< 105
MA; score	39	1	271	1
Variables in 24 h urine
Sodium; mM	182 [186]	42	390	< 230
Calcium; mM	5.4	1.1	17	< 7.5
Pi; mM	30	9.1	61	< 39
Potassium; mM	62 [150]	15	198	40-90
Mg; mM	4.1 [281]	1.3	9.2	4-6
Urea-nitrogen; mM	0.78	0.15	1.85	< 2.5
Oxalate; mM	0.26 [76]	0.08	0.45	< 0.51
pH	5.94	4.82	7.20	< 7.0
Fasting blood and plasma
Blood bicarbonate; mM/l	23.6 [276]	12	31	> 24
Blood pH	7.40 [276]	7.32	7.49	> 7.35
Creatinine; mM/l	0.09	0.06	0.12	< 0.12
Total Mg; mM/l	0.85	0.65	1.0	0.73-1.0
Ultrafiltrable Mg; mM/l	0.66	0.50	0.91	< 0.92
Pi; mM/l	0.95	0.60	1.59	> 1.13
Glucose; mM/l	4.9 [258]	3.6	6.8	< 4.7
Insulin; pM/l	115 [258]	72	646	< 108
Fasting urine (2 h)
Volume; ml	228	48	760	< 400
pH	6.19	4.41	7.60	4.8 - 6.6
Mg; mM	0.26	0.05	0.79	> 0.17
FE Mg;%	3.2	1.0	5.0	1.9-3.8*
Potassium; mM	8.0	1.8	36	< 12
Ca; mM	0.33	0.03	1.3	< 0.43
Pi; mM	1.2	0.1	3.9	< 2.4
Sodium; mM	13	2	47	< 20
Glucose; µM	483	28	3017	< 555
Citrate; µM	275 [244]	42	741	> 212
Oxalate; µM	24 [273]	1	45	< 34
Urea-nitrogen; µM	71	5	407	< 80
Total protein; mg	8.4 [262]	0.8	171	< 4
Albumin; mg	1.7 [191]	0.05	57	< 0.6
N-Alb-P; mg	6.2 [191]	0